Tonic sound-evoked motility of cochlear outer hair cells in mice with impaired mechanotransduction

Abstract

Cochlear outer hair cells (OHCs) transduce sound-induced vibrations of their stereociliary bundles into receptor potentials that drive changes in cell length. While fast, phasic OHC length changes are thought to underlie an amplification process required for sensitive hearing, OHCs also exhibit large tonic length changes. The origins and functional significance of this tonic motility are unclear. Here, in vivo cochlear vibration measurements reveal tonic, sound-induced OHC motility in mice with stereociliary defects that impair mechanotransduction and eliminate cochlear amplification. Tonic motility in impaired mice was physiologically vulnerable but weakly related to any residual phasic motility, possibly suggesting a dissociation between the underlying mechanisms. Nevertheless, a simple model demonstrates how tonic responses in both normal and impaired mice can result from asymmetric mechanotransduction currents and be large even when phasic motility is undetectable. Tonic OHC responses are therefore not unique to sensitive ears, though their potential functional role remains uncertain.

Keywords: cochlear mechanics; mechanotransduction; optical coherence tomography; outer hair cell motility.

Figure 1.. Theoretical dependence of tonic OHC responses on asymmetric mechanotransduction.

(a) Schematic of the top of an OHC and its stereociliary bundle. The stereocilia are anchored to the apical OHC surface by stiff rootlets. Deflection of the bundle stretches the tip links that connect adjacent stereociliary rows, thus opening ion channels. (b) First-order Boltzmann function approximating the relationship between bundle displacement and the resulting mechanotransduction current. If the bundle’s resting position is biased away from the center of the function, sinusoidal bundle displacements will produce asymmetric currents that contain a tonic (i.e., direct current or 0 Hz) component. Low-pass filtering the current waveform allows direct visualization of the tonic component (gray curve).

Figure 2.. Sound evokes tonic OHC contractions in the apex of the WT mouse cochlea.

(a) Cross-sectional OCT image of an apical cochlear location tuned to ~9 kHz in a WT C57BL/6J mouse. The three fluid-filled scalae (SV = scala vestibuli, SM = scala media, ST = scala tympani) are indicated along with Reissner’s membrane (RM), the tectorial membrane (TM), and basilar membrane (BM). Scale bar = 100 μm. (b) Magnified view of the organ of Corti indicating the approximate location of the OHCs (dotted white lines), which was inferred from the positions of the tunnel of Corti (ToC) and TM. Stars indicate measurement points where tonic displacements were measured (RL = reticular lamina, LC = lateral cell region, OHC-DC = OHC-Deiters’ cell junction). Scale bar = 50 μm. (c) Organ of Corti schematic with relevant anatomy labeled (IHC = inner hair cell, DC = Deiters’ cell). (d) Average BM displacement magnitudes and phases as a function of stimulus frequency and sound pressure level (SPL) in WT mice. Displacement magnitudes are normalized to the evoking stimulus pressure (in Pascals, Pa) to highlight the nonlinearity exhibited by the responses. Responses to 10 and 20 dB SPL tones were also obtained but are not shown, for clarity. Dashed-dotted lines indicate ± 1 SE, though these are generally obscured by the average curves. (e-f) Displacement waveforms measured from the RL and OHC-DC junction for a 9 kHz, 60 dB SPL stimulus (e) and a 5 kHz, 90 dB SPL stimulus (f) in an individual mouse. Low-pass filtered waveforms (gray curves) reveal that the RL and OHC-DC junction tonically moved toward one another during the stimulus. Different displacement scales are used in e and f, and stimulus pressure waveforms are arbitrarily scaled. (g) Tonic displacements measured from the OHC region and surrounding structures for a 5 kHz, 90 dB SPL stimulus in WT mice (n = 18 for the OHC-DC junction; n = 17 for all other points). Positive and negative displacements are toward SV and ST, respectively, and are plotted on a linear scale. (h) Average tonic displacements of the OHC-DC junction vs. stimulus level at four frequencies (n = 5, 13, 6, and 6 mice for 3, 5, 7, and 9 kHz stimuli, respectively). Displacements are plotted on a logarithmic scale and the dashed line illustrates linear growth. Error bars indicate ± 1 SE.

Figure 3.. Sound evokes tonic OHC contractions in mice with impaired mechanotransduction and minimal nonlinear BM amplification.

(a-b) Average BM displacement gain (re stimulus pressure) and phase as a function of stimulus frequency and level for salsa (a) and Triobp (b) mice. Displacement gain curves for 50–100 dB SPL stimuli are indistinguishable, indicating the linearity of the responses and the absence of BM amplification. (c-f) Displacement waveforms of the RL and OHC-DC junction in individual salsa (c,e) and Triobp (d,f) mice for a 5 kHz, 90 dB SPL stimulus. While tonic displacements could not be detected in some mice (c-d), they were present in many others (e-f; gray curves are the low-pass filtered waveforms). Stimulus waveforms are arbitrarily scaled. See Supplementary Fig. S3 for more examples of tonic displacements in impaired mice. (g-h) Tonic displacements measured from different points within the OHC region and surrounding structures for 90 dB SPL stimuli in salsa (g) and Triobp (h) mice. Positive and negative displacements are toward SV and ST, respectively. Stimulus frequencies were set within 0.5 kHz of the frequency eliciting maximum BM displacement in each mouse and ranged from 4 to 5.5 kHz. (i) Average absolute tonic displacement magnitudes for the OHC-DC junction and RL compared across strains (individual data are shown in g and h and Fig. 2g). The number of mice included in each average is shown in parentheses and error bars indicate 95% confidence intervals. Tonic displacements were compared across strains using one-way ANOVAs (for OHC-DC data, F_2,91 = 37.58, p < 0.005; for RL data, F_2,89 = 31.93, p < 0.005). Asterisks indicate statistical significance of post-hoc comparisons with Bonferroni corrections (** p < 0.005, *** p < 0.0005, ns = not significant).

Figure 4.. Tonic displacements in impaired mice are diminished after death, salicylate administration, and aging.

(a) Waveforms of RL and OHC-DC junction displacements for a 4 kHz, 90 dB SPL stimulus in an individual Triobp mouse before and ~40 min after death. Low-pass filtered waveforms (gray curves) show the tonic displacements, which were largely reduced postmortem. (b) Average absolute tonic displacements of the OHC-DC junction before and after death in WT, salsa, and Triobp mice. Averages only include data from mice with absolute tonic displacements ≥ 1.5 nm in the live condition, and for which postmortem measurements were obtained < 45 min after death. Error bars indicate 95% confidence intervals and individual data are shown with semi-transparent symbols. Tonic displacements were significantly reduced postmortem in all strains, as assessed by paired t-test (for WT, t₁₁ = 18.38, p < 0.0005; for salsa, t₇ = 3.60, p = 0.009; for Triobp, t₂₃ = 4.85, p < 0.0005). (c) Waveforms of RL and OHC-DC junction displacements for a 5 kHz, 90 dB SPL stimulus in an individual Triobp mouse before and ~20 min after administration of salicylate to the round window membrane. (d) Absolute tonic displacements of the OHC-DC junction for a 90 dB SPL stimulus before and < 30 min after salicylate administration in two WT mice, one salsa mouse, and two Triobp mice (responses in a WT and salsa mouse are shown in Supplementary Fig. S4). The stimulus frequency was 4.5 kHz for one Triobp mouse and 5 kHz for all other mice. Lines connect data from each mouse. (e-f) Absolute tonic OHC-DC displacements for 90 dB SPL stimuli as a function of age in salsa (e) and Triobp mice (f). Spearman’s rank correlations (ρ) and R² values are provided in each panel, and dashed lines indicate linear fits. In b, e, and f, asterisks indicate statistical significance (* p < 0.05, ** p < 0.005, *** p < 0.0005).

Figure 5.. Nonlinear, cycle-by-cycle OHC electromotility is present in some impaired mice but is not strongly correlated with tonic displacements.

(a-b) Average displacement gains (a) and phases (b) for the OHC-DC junction and TM as a function of frequency and stimulus level in WT mice (n = 17). Average phase differences between the TM and OHC-DC junction are shown in the bottom panel of b. Gains and phase differences are shown for stimulus levels of 30 to 90 dB SPL in 10 dB steps, while phases are shown only for 80 dB SPL stimuli, for clarity. Responses to higher levels are shown with darker/thicker lines. Dashed-dotted lines indicate ± 1 SE. Nonlinear gain was observed down to at least 2 kHz, and roughly out-of-phase motion was observed at low frequencies, even at high stimulus levels. (c-d) As in a-b, but for an individual salsa mouse. Gains and phase differences are shown for stimulus levels of 70 to 100 dB SPL in 10 dB steps, and phases are shown only for 80 dB SPL stimuli. Gain curves largely overlap, indicating response linearity, and phase differences are close to 0. (e-f) As in c-d, but for an individual Triobp mouse whose responses exhibited a small amount of nonlinearity and modest phase differences. Such nonlinearity and phase differences disappeared postmortem, as shown in Supplementary Fig. S5. (g) Absolute tonic OHC-DC displacement magnitudes plotted vs. the nonlinear gain observed in OHC-DC displacements for all salsa and Triobp mice. Nonlinear gain was quantified by taking the average difference between displacement gains at 70 and 100 dB SPL for stimulus frequencies of 1 to 4 kHz. (h) Absolute tonic OHC-DC displacement magnitudes plotted vs. the phase difference between TM and OHC-DC displacements for all salsa and Triobp mice. Phase differences were calculated for 80 dB SPL stimuli and averaged for stimulus frequencies of 1 to 4 kHz. In g-h, Spearman’s rank correlations (ρ) and R² values are provided, with asterisks indicating significant correlations (* p < 0.05, ** p < 0.005, *** p < 0.0005, ns = not significant). Dashed lines are linear fits. Similarly weak-to-modest correlations were found when comparing nonlinearity observed in TM motion to tonic RL displacements, as shown in Supplementary Fig. S6.

Figure 6.. DPOAEs are present at high stimulus levels in some impaired mice and are modestly correlated with tonic displacement magnitudes.

(a) DPOAE amplitudes in WT, salsa, and Triobp mice for 85 dB SPL stimuli plotted as a function of the f₂ stimulus frequency (f₂/f₁ = 1.22). Data for all individual mice are overlaid and dotted portions of the curves indicate data not meeting the signal-to-noise criterion. (b) Absolute tonic OHC-DC displacement magnitudes plotted vs. DPOAE amplitudes for all salsa and Triobp mice. For each mouse, DPOAE amplitudes were averaged across f₂ frequencies falling within ± 1 kHz of the frequency that elicited maximum BM displacement. Spearman’s rank correlations (ρ) and R² values are provided, with significant correlations indicated by asterisks (* p < 0.05, ** p < 0.005, *** p < 0.0005). Dashed lines are linear fits. The physiological origin of the measured DPOAEs was confirmed by postmortem comparisons, as shown in Supplementary Fig. S7.

Figure 7.. Changes in the mechanotransducer function can account for the presence of tonic responses in the absence of significant nonlinear amplification.

(a) Boltzmann functions used to represent the relationship between OHC stereociliary bundle displacement and mechanotransduction current in WT, salsa, and Triobp mice. Waveforms are shown for a 5 kHz, 30 nm input and the corresponding outputs for each strain. Tonic components in the output waveforms are shown with gray curves. For all mice, the Boltzmann slope parameter a was 0.17 nm⁻¹. In WT, salsa, and Triobp mice, respectively, the maximum current ($I_{\mathrm{m}\mathrm{a}\mathrm{x}}$) was 1, 0.16, and 0.4 and the operating point (x₀) was 3.8, 18, and 19 nm. Active OHC region displacements were approximated by low-pass filtering the Boltzmann output and scaling by a factor of 125. To model the total OHC region displacement, the active response waveforms were summed with an estimate of the passive displacement, which was taken as the underlying BM displacement (i.e., the Boltzmann’s input) scaled by a factor of 0.56. (b) Average phasic and tonic displacements of the OHC-DC junction for 5 kHz tones varied in level (symbols) compared with modeled displacements (lines) in WT, salsa, and Triobp mice. Averages only include data meeting the signal-to-noise criterion from mice with tonic responses greater than 1.5 nm at 90 dB SPL and with BM displacements obtained in 5 dB steps over the same range of stimulus levels. Averages are shown when such clean data were obtained in at least 60% of mice. Responses from all individual mice, including those not meeting the above criteria, are shown in Supplementary Fig. S8. Dashed lines indicate linear growth.